Oct 16, 2012 · This paper explores a new framework for reinforcement learning based on online convex optimization, in particular mirror descent and related algorithms.
This paper explores a new framework for rein- forcement learning based on online convex opti- mization, in particular mirror descent and related algorithms.
This paper explores a new framework for reinforcement learning based on online convex optimization, in particular mirror descent and related algorithms.
A new framework for reinforcement learning based on online convex optimization, in particular mirror descent and related algorithms, and a new family of ...
This paper explores a new framework for reinforcement learning based on online convex optimization, in particular mirror descent and related algorithms.
This paper explores a new framework for reinforcement learning based on online convex optimization, in particular mirror descent and related algorithms.
Sparse q-learning with mirror descent. In Proceedings of the Twenty-. Eighth Conference on Uncertainty in Artificial In- telligence, Catalina Island, CA, USA ...
We propose two new algorithms for the sparse re- inforcement learning problem based on different formulations. The first algorithm is an off-line.
We analyze continuous-time mirror descent applied to sparse phase retrieval, which is the problem of recovering sparse signals from a set of magnitude-only ...
People also ask
Why use mirror descent?
Does deep Q learning use gradient descent?
In comparison to Lan (2022), our work sets forth a different framework to analyze mirror-descent type algorithms for regularized policy optimization,.